Search Results (306)

Accurate electricity price forecasting (EPF) is essential for market participants to optimize trading strategies and for power systems to accommodate the increasing penetration of volatile renewable energy sources. However, electricity price series are characterized by strong nonlinearity, high volatility, and significant structural breaks, which pose substantial challenges to conventional forecasting models. Although numerous hybrid deep learning models have been proposed for EPF, most existing approaches either overlook structural breaks or treat them as outliers rather than as signals of regime shifts, often resulting in systematic forecasting degradation when market conditions change abruptly. To address this issue, this study proposes COCAL-TTL, a novel multi-stage structural break-aware forecasting framework that integrates regime-adaptive data partitioning with a functionally differentiated hybrid deep learning architecture. First, a joint detection scheme combining the Iterated Cumulative Sum of Squares (ICSS) algorithm and the Chow test is employed to partition Spanish electricity market data from 2014 to 2023 into distinct regimes. Within each regime, CEEMDAN is applied to extract multi-scale features, which are subsequently reconstructed into trend, periodic, and random components based on an independent sample t-test and Fast Fourier Transform (FFT). The CNN-SE Attention-LSTM (CAL) model, with hyperparameters optimized by the Osprey Optimization Algorithm (OOA), serves as the primary forecasting engine. In addition, a dedicated heterogeneous error correction module, namely TTL, is introduced, in which Temporal Convolutional Network, Transformer, and LSTM are designed to capture local transients, long-range dependencies, and transitional dynamics in the residual series, respectively. Empirical results demonstrate that compared with the Naive benchmark, COCAL-TTL achieves percentage MAPE improvements of 58.48% and 48.97% in low- and high-volatility regimes, respectively. These findings indicate that the proposed structural break-aware framework provides a robust data-driven solution for EPF under heterogeneous market conditions and offers technical support for stable electricity market operation in the context of renewable energy integration. Full article

Efficient multi-hop prediction over large-scale spatio-temporal knowledge graphs of aircraft taxiing trajectories remains challenging, as existing methods focus either on static multi-hop relations or on accuracy improvement for spatio-temporal single-hop predictions, leading to computational inefficiency. This paper proposes a vector-index-supported multi-hop prediction method. First, a knowledge graph embedding technique that integrates spatio-temporal features maps the trajectory graph into a low-dimensional complex vector space. Then, a hierarchical query acceleration structure based on IndexIVFFlat is constructed. A clustering strategy guided by the distribution of trajectory data partitions the vector space into subspaces, and approximate nearest neighbor search within those subspaces rapidly prunes the candidate set to accelerate multi-hop retrieval. Experiments on real aircraft taxiing trajectory datasets and general benchmarks show that the proposed method substantially improves prediction efficiency while maintaining competitive accuracy. The results demonstrate that the vector index mechanism effectively balances accuracy and efficiency, and the efficiency has been improved by at least 56.65%. This work provides a key technical foundation for real-time analysis and intelligent prediction of large-scale aircraft taxiing trajectories. Full article

Mountain stream ecosystems are often considered among the least disturbed freshwater environments; however, increasing land-use pressures may affect their ecological integrity even under apparently high-water quality conditions. This study aimed to assess the relative influence of landscape, physicochemical, and hydromorphological factors on benthic macroinvertebrate communities in three sub-catchments (Ambroz, Jerte, and Tiétar) of the Sierra de Gredos (Central Spain). A total of 33 sampling sites were surveyed, and macroinvertebrate assemblages were analyzed in relation to environmental variables using partial Redundancy Analysis (pRDA) and variance partitioning. All sites were classified as having “Excellent” ecological status based on the Iberian Biological Monitoring Working Party (IBMWP) index. However, multivariate analyses revealed clear spatial patterns and responses to environmental gradients. Results indicated that catchment-scale landscape characteristics defined the pool of potential colonizers, while local physicochemical and hydromorphological conditions acted as secondary filters structuring macroinvertebrate assemblages. Landscape variables explained the largest fraction of variance in community structure (30.6%), followed by physicochemical parameters (29.0%) and hydromorphological indices (24.9%), with a significant shared component (16.5%) indicating interactions among drivers. Agricultural land use, particularly in the Jerte sub-catchment, was associated with shifts in community composition, favoring tolerant taxa such as Diptera, while sub-catchments dominated by natural vegetation supported higher richness of sensitive groups, including Ephemeroptera and Plecoptera. These findings highlight the importance of multi-scale processes in structuring mountain stream communities and reveal limitations of traditional biotic indices in detecting early ecological changes. The results support the integration of catchment-scale variables into ecological assessment frameworks and emphasize the need for preventive, basin-scale management strategies to maintain ecological integrity under increasing anthropogenic pressure. Full article

Existing single-image super-resolution methods for remote sensing images suffer from insufficient global receptive fields, weak high-frequency texture recovery, and excessive computational complexity. To address these issues, this paper proposes DFSMamba, a novel spatial–frequency collaborative modeling framework. First, Semantic Continuous-Sparse Attention enhances semantic perception through dynamic chunking and sparse connections while maintaining linear complexity, effectively alleviating the semantic truncation problem caused by fixed window partitioning. Second, the Adaptive State-Space Module employs parallel forward and backward state-space model branches to achieve bidirectional long-range dependency modeling and introduces an activation-guided feature fusion mechanism to adaptively enhance semantically relevant regions. Third, the Discrete Fourier Transform Module maps images to the frequency domain, establishes a global lossless receptive field, and explicitly enhances high-frequency details, compensating for the insufficient utilization of frequency-domain information in pure spatial-domain methods. Experiments on five public datasets demonstrate that DFSMamba outperforms mainstream CNN, Transformer, and Mamba-based methods across ×2 to ×4 scales. On the AID×3 task, it achieves a PSNR of 31.48 dB, exceeding MambaIRv2 by 1.07 dB. Ablation studies verify the positive synergistic effect of the three modules, with the full configuration achieving a PSNR improvement of 0.85 dB over the single-module setup. Fine-grained category, multi-scale input, and loss function experiments further confirm its robustness and generalization capability, particularly in edge and texture detail reconstruction. Full article

As the offshore wind industry scales toward 15 MW capacity, the reliability of Direct-Drive Permanent Magnet Synchronous Generators (DD-PMSGs) becomes critical. However, real-world run-to-failure data for these massive, multi-pole machines is virtually non-existent, creating a barrier for developing effective data-driven diagnostic systems. This study proposes a high-fidelity framework for detecting permanent magnet faults in the International Energy Agency (IEA) 15 MW Reference Wind Turbine. Using Finite Element Analysis (FEA), a dataset (magnetic flux and back electromotive-force (EMF)) capturing the electromagnetic signatures of healthy and faulty states of a PMSG under varying severities is generated. To improve the power of computer vision, 1D time-series signals were transformed into 2D images. Specifically, Gramian Angular Fields (GAFs) and Recurrence Plots (RPs) were applied to magnetic flux density signals, while Markov Transition Fields (MTFs) were applied to back-EMF signals. These representations were then fused into multi-channel Red-Green-Blue (RGB) images and processed via a ResNet-18 Deep Transfer Learning model using a strictly non-overlapping, leakage-free dataset partitioning strategy. The proposed framework achieved a classification accuracy of 99.45% on noise-free data. Furthermore, robustness testing under varying levels of Additive White Gaussian Noise (AWGN) (30 dB, 40 dB, and 50 dB Signal-to-Noise Ratio (SNR)) demonstrated sustained high performance, maintaining over 90% accuracy even under severe 30 dB noise conditions. Comparative analysis proved that this multi-channel fusion significantly outperforms single-channel encoding methods, which collapse under heavy noise, validating the scalability of the framework and applicability for next-generation condition monitoring in harsh offshore environments. Full article

Magnitude alone fails to capture the full complexity of pavement roughness; its spatial distribution along a road is equally vital for effective maintenance planning. While traditional assessment has long relied on specialized survey vehicles, the rise of mobile crowdsensing now allows for massive data acquisition via smartphone sensors. This study investigates the spatial structure of pavement roughness using crowdsensed data from the SmartRoadSense platform. Roughness is quantified through the Power of Prediction Error (PPE) indicator derived from smartphone accelerometer signals. The dataset consists of 475 observations sampled at 20 m intervals over approximately 9.5 km of the A3/E45 motorway in southern Italy. A multi-scale spatial–statistical framework is adopted to analyse the roughness signal. The analysis includes the evaluation of scale-dependent statistical descriptors (mean and coefficient of variation), as well as spatial correlation, spectral, and entropy-based measures. The results indicate a short spatial correlation length (approximately 60–100 m) and the absence of a dominant spatial wavelength, suggesting that pavement roughness behaves as a localized multiscale process. A complementary segmentation analysis based on Classification and Regression Trees (CART) is performed to explore the spatial partitioning of the roughness signal. Our analysis indicates that segmentation complexity spikes once the minimum node size drops below roughly 10 observations. This trend points to the existence of localized irregularities that coarser scales simply overlook. Ultimately, these results suggest that mean roughness values alone are insufficient for describing pavement condition and that hybrid spatial–statistical approaches may support more scalable, data-driven, and spatially targeted pavement monitoring strategies for sustainable transportation infrastructure management. Full article

Malicious URLs exhibit diverse patterns and rapid evolution in complex network environments. Traditional detection methods often focus solely on local patterns or global dependencies, failing to capture multi-scale information effectively, thus limiting their performance on obfuscated characters and structural variants. To address these limitations, this paper proposes PLG-URLNet, a malicious URL detection method based on layer-by-layer interaction of local and global features. The model adopts a two-branch structure: a convolutional branch extracts local patterns and contextual information, while a Transformer branch models global dependencies and long-range relationships. A hierarchical interaction module (PLGIM) embedded in each layer enables effective feature integration, enhancing representation capabilities progressively. A lightweight decoder-type fusion layer is further introduced for final discrimination. To avoid over-estimating generalization ability caused by random train–test splitting, both standard evaluation protocols and time-sequential partitioning are adopted. Experiments show that PLG-URLNet achieves an accuracy of 98.59%, maintaining performance advantages in stringent temporal generalization scenarios and effectively identifying obfuscated and variant URLs. This method provides a feasible solution for malicious URL detection. Full article

Paclitaxel (Taxol), a taxane diterpenoid from Taxus species, is a clinically important microtubule-stabilizing anticancer agent widely used in chemotherapy. However, its supply remains limited by precursor scarcity and the molecule’s structural complexity. The biosynthetic pathway from geranylgeranyl diphosphate (GGPP) to paclitaxel is estimated to involve 19 to 23 enzymatic steps. Recent multi-omics approaches have substantially elucidated this pathway, yet key mechanistic questions persist, notably the formation of the oxetane ring. Complete heterologous biosynthesis is further hampered by poor cytochrome P450 (CYP) expression in non-native hosts and insufficient metabolic flux. This review synthesizes advances across four themes: (1) progressive elucidation of the biosynthetic pathway, with emphasis on the CYP-mediated oxygenation cascade and oxetane ring formation; (2) genomic and regulatory insights from Taxus genome assemblies, transcription factor networks, and spatial multi-omics; (3) metabolic engineering in microbial hosts, including Escherichia coli, Saccharomyces cerevisiae, and non-conventional chassis; and (4) plant-based heterologous production platforms. Critical bottlenecks are identified, including unresolved enzymatic steps, CYP functional expression, flux partitioning, and bioprocess scale-up. Strategies to overcome these challenges are discussed. Full article

Accurate prediction of loan defaults is essential for financial institutions seeking to minimize credit losses and maintain portfolio stability. In the vehicle financing segment of emerging markets, real-world datasets frequently exhibit extreme class imbalance ratios that far exceed those encountered in standard benchmark corpora, posing severe challenges for conventional machine learning pipelines. This study introduces a gradient boosting framework integrating Weight of Evidence (WoE) transformation, Bayesian hyperparameter optimization, and three complementary classifiers—Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), and Categorical Boosting (CatBoost)—to predict vehicle loan default risk. The methodology is evaluated on a large-scale, fully anonymized Turkish vehicle loan dataset ($N=207,572$) with an extreme imbalance ratio of 1:1133 (183 defaults versus 207,389 non-defaults). A strict three-way data partition (60% training, 20% validation, 20% test) is adopted to ensure leakage-free model selection and unbiased performance estimation. A multi-stage experimental pipeline is developed encompassing: (i) statistical feature selection via Mann–Whitney U and chi-square tests with adaptive thresholding, (ii) a comparative analysis of seven resampling strategies including Synthetic Minority Oversampling Technique (SMOTE) variants, Adaptive Synthetic Sampling (ADASYN), and focal loss weighting, (iii) a greedy forward selection ensemble procedure for heterogeneous model fusion, and (iv) a systematic training-set size sensitivity analysis across eight majority undersampling ratios. Under the leakage-free evaluation protocol, the highest-AUC individual model (LightGBM with SMOTE-ENN) achieves an Area Under the Curve (AUC) Receiver Operating Characteristic (ROC) of 0.710 (95% bootstrap CI: 0.614–0.798), while CatBoost with cost-sensitive weighting exhibits superior operational metrics (KS $=0.389$, PR-AUC $=0.011$). The greedy ensemble procedure exhibits high selection instability with only 37 validation-set positives, providing a methodological finding on the minimum sample requirements for reliable ensemble construction under extreme scarcity. Ablation results confirm that WoE encoding contributes 3.1 percentage points to the overall AUC gain. Tree SHAP-based interpretability analysis identifies the financing-to-age ratio, WoE-encoded occupation group, and log financing amount as the primary predictive drivers, with cross-model stability confirmed via Spearman rank correlation. A decision support analysis provides precision–recall curves, a Brier score of 0.0082, reliability diagrams, and threshold-dependent performance at operationally plausible review rates. Fairness evaluation across gender and marital status subgroups demonstrates that threshold-dependent metrics such as Disparate Impact Ratio and Equalized Odds Gap are inherently compromised under extreme minority scarcity, whereas rank-based subgroup AUC analysis with bootstrap 95% confidence intervals preserves meaningful discriminative assessment. These findings provide an empirically validated framework for credit default prediction in highly imbalanced and data-scarce financial environments. Full article

Heterostructured metals and alloys are designed with spatial variations in strength and hardening that produce synergy beyond the rule of mixtures. This review surveys face-centered cubic (FCC), body-centered cubic (BCC), and hexagonal close-packed (HCP) systems, including architectures formed or modified by rolling and related severe plastic deformation routes, and examines them under tension, compression, and shear. Across material classes, mechanical incompatibility between hetero-zones drives stress partitioning and plastic strain gradients that store geometrically necessary dislocations near zone boundaries. The associated internal back and forward stresses sustain work hardening, delay instability, and influence localization and damage initiation. We evaluate continuum, crystal plasticity, dislocation-based mesoscale, and atomistic approaches by whether they predict these internal fields and whether they are validated against internal-field measurements. Key observations are that predictive models require physically identifiable intrinsic length scales, experimentally constrained interface laws, and careful separation of mechanisms to avoid double-counting when gradient and kinematic terms coexist. Major gaps remain in parameter identifiability for multi-zone and nonlocal formulations, in transferability across processing routes and loading modes, and in community benchmarks that couple well-characterized microstructures with multimodal measurements. Recommendations are provided for validation targets and benchmark campaigns to accelerate predictive design. Full article

This paper applies physics-informed neural networks (PINNs) to laterally excited liquid sloshing in a two-dimensional rectangular tank, where near-resonant forcing (${\omega }_e/{\omega }_1=0.9$) produces a multi-frequency beating response with a period of approximately $10T_1$. Linearized potential flow theory governs the problem; the network learns the velocity potential $\phi (x,z,t)$ while the free-surface elevation $\eta$ is injected analytically. Two training obstacles specific to forced sloshing are analyzed. First, a zero-solution trap arises because the trivial solution $\widehat{\phi }=0$ satisfies all equations except the free-surface conditions, whose residuals are roughly ${10}^{-4}$ times smaller than the Laplace residual; characteristic-scale normalization combined with loss weighting (${\lambda }_D={\lambda }_K=100$) breaks this trap. Second, spectral bias prevents standard MLPs from resolving the three co-existing frequencies (${\omega }_1$, ${\omega }_e$, $\mathsf{\Delta }\omega$); a Fourier time embedding that augments the input from 3 to 9 dimensions overcomes this limitation. Two additional techniques further reduce errors: a hard-wall boundary condition enforced exactly via a $\cos (\pi x/B)$ spatial embedding, which eliminates wall collocation points; and a gradient-enhanced Laplace regularizer (${\parallel \nabla \left({\nabla }^2\widehat{\phi }\right)\parallel }^2$) that constrains velocity smoothness through third-order automatic differentiation. An ablation study shows that these four techniques progressively reduce the horizontal velocity error from ${\epsilon }_u=12.46\%$ to $0.84\%$. Results are validated against a viscous finite-difference benchmark. Over one beating cycle the errors are ${\epsilon }_{\eta }=0.15\%$, ${\epsilon }_u=0.84\%$, and ${\epsilon }_w=1.65\%$. A frequency parameter study across ${\omega }_e/{\omega }_1$ = 0.5–1.1 gives ${\epsilon }_{\eta }<0.25\%$ and ${\epsilon }_u<2.3\%$ for all near-resonance cases. For long-time simulation, a time-domain decomposition strategy with transfer learning partitions the domain into one-beat windows; extending to five beating cycles ($50T_1$) yields ${\epsilon }_u=3.43\%$ and ${\epsilon }_{\eta }=0.30\%$ with no monotonic error accumulation across windows. The methodology is then extended to a three-dimensional rectangular tank ($B\times W\times H$) with bi-directional lateral excitation. The 3-D formulation introduces the y-dimension into the Laplace equation (${\nabla }^2\phi ={\phi }_{xx}+{\phi }_{yy}+{\phi }_{zz}=0$), adds transverse wall boundary conditions ($\partial \phi /\partial y=0$) enforced exactly via a $\cos (\pi y/W)$ embedding, and extends the Fourier time embedding from 9 to 16 dimensions to accommodate six physical frequencies. The bi-directional excitation excites both $(m,0)$ and $(0,n)$ modal families, producing a genuinely three-dimensional beating response. Experimental results verify that the proposed methods can be well generalized to three-dimensional scenarios. Within a single beating cycle, the relative errors reach ${\epsilon }_{\eta }=0.24\%$, ${\epsilon }_u=1.31\%$, ${\epsilon }_v=1.78\%$ and ${\epsilon }_w=2.32\%$, with a total training time of 2499 s. By applying time domain decomposition to carry out two-cycle three-dimensional simulations, the model can steadily maintain satisfactory prediction precision across segmented time intervals, achieving overall errors of ${\epsilon }_{\eta }=0.30\%$ and ${\epsilon }_u=1.32\%$. Full article

The rapid growth of maritime communication networks introduces significant challenges to routing optimization, arising from large-scale network topologies, highly dynamic node mobility, and stringent real-time communication requirements. Conventional routing algorithms often exhibit limited scalability and poor adaptability when facing frequent topology variations. The routing problem is modeled as a multi-agent distributed decision-making process, where each node acts as an autonomous agent. In this paper, we propose a graph-partitioned hierarchical graph representation learning framework (GP-HGRL) for scalable and continual routing optimization in dynamic maritime networks. By explicitly modeling the network as a time-evolving graph, GP-HGRL first partitions the global topology into topology-aware subgraphs, enabling distributed learning and inference with reduced computational complexity. A hierarchical graph neural network architecture is then developed to jointly capture intra-subgraph local structures and inter-subgraph global dependencies, producing topology-aware embeddings for routing decision-making. Based on the learned representations, a deep reinforcement learning policy is employed to perform distributed next-hop routing decisions. To effectively handle topology dynamics induced by node mobility and link variations, we further introduce a continual graph learning mechanism that selectively updates representations and routing policies only within affected subgraphs, thereby avoiding costly global retraining and preserving routing stability. Extensive simulations demonstrate that GP-HGRL consistently outperforms shortest-path routing and existing reinforcement learning-based approaches in terms of packet delivery ratio, retransmission rate, packet loss, and training efficiency under various network loads and dynamic conditions. Full article

This study investigates the physiological trade-off between biomass yield and sugar concentration in five sweet sorghum genotypes to evaluate how carbon partitioning influences bioethanol potential. Field experiments were conducted over the 2019–2020 seasons in the East Marmara transitional zone of Türkiye, under irrigated and rain-fed regimes. Results revealed a highly significant genotype × water regime interaction (p < 0.001). A distinct trade-off was identified: while the hybrid ‘Teide’ maximized juice volume under irrigation (2427.67 L ha⁻¹), ‘Leoti’ maintained superior sugar stability (18.38 °Brix) under moisture deficit. Genotype plus Genotype × Environment Interaction (GGE) biplot analysis indicated that ‘Early Sumac’ provided the highest environmental buffering, balancing productivity and sugar density across water regimes. Principal Component Analysis (PCA) demonstrated that plant height (averaging 214.2 cm) was positively associated with juice yield and concentration. Under irrigation, ‘Teide’ produced the highest bioethanol yield (1690.7 L ha⁻¹), whereas ‘Nutrihang’ led output under rain-fed conditions. While these site-specific trends offer valuable insights into local bioenergy stability, further multi-location trials are necessary to confirm these patterns on a broader scale. The findings conclude that feedstock selection must be categorized by water availability to optimize sweet sorghum-based bioenergy systems in water-limited environments. Full article

Salt-bearing foreland fold–thrust belts represent a critical tectonic system for ultra-deep hydrocarbon exploration. In the Kalasu structural belt of the Kuqa Depression—characterized by the “four extremes” of ultra-high temperature, pressure, salinity, and stress—conventional single-detachment models fail to adequately resolve the complex subsalt structures. To address this challenge, this study integrates high-resolution 3D seismic data, field outcrop observations, well logs, balanced cross-sections, and particle image velocimetry (PIV)-monitored physical modeling to propose a ramp–flat multi-detachment model. Our results demonstrate that deformation is governed by four regional detachment horizons: gypsum-salt layers, thick mudstones, coal-bearing strata, and the basement, which vertically partition the basin into six tectonic units: supra-salt, salt, subsalt, supra-coal, coal, and sub-coal basement. The structural architecture is controlled by five key factors: (1) paleo-uplift geometry, (2) distance from the South Tianshan orogenic front, (3) orientation of basin-bounding faults, (4) regional stress regime (pure compression versus transpression), and (5) rheological contrasts among detachment layers. The kinematic evolution follows a progressive sequence: basement-involved thrusting → multi-level ramp–flat detachment folding → cover detachment. Three primary trap levels are identified—subsalt, supra-coal, and sub-coal—hosting six distinct trap styles: pop-up anticlines, imbricate faulted anticlines, structural triangle zones, fault-bend fold anticlines, supra-coal anticlines, and inter-coal/sub-coal anticlines. Notably, under transpressional stress, oblique paleo-uplifts control the formation of enigmatic “fish-scale” arcuate trap belts composed of fault-bend fold anticlines. Full article

Trans-border data circulation across multi-jurisdictional boundaries faces an operational conflict between ownership provenance prerequisites and data minimisation mandates, compounded by the tight coupling of large data payloads with synchronous state consensus ledgers, which forces replication of feature matrices across all consensus nodes and leads to network saturation. Existing frameworks remain unequipped to resolve this, as coupling in-band payload routing with synchronous state ledgers generates communication overheads scaling with data volume. The proposed Trusted Data Space with Registration (TDSR) implements a four-layer protocol stack. A dual-plane topology establishes a decoupled storage–ledger mechanism, partitioning asynchronous payload datastores and synchronous consensus ledgers to sustain throughput independent of data dimensionality. Navigating this infrastructure, the Unified Data Resource Identifier (UDRI) executes out-of-band cross-domain routing without exposing verifier intents. Driven by the Oblivious Data Asset Registration (ODAR) mechanism, a two-phase, four-algorithm lifecycle dictates end-to-end ownership provenance. This execution shifts hypothesis testing to isolated sandboxes via an algorithm-agnostic mathematical contract, capping external data transit at a constant leakage bound. A deployed testbed across the Guangdong-Hong Kong-Macao Greater Bay Area validates the proposed architecture, supporting data circulation across divergent legal jurisdictions. Full article